Upayan Bakshi

Key Skills

Software

CVAT

Labelbox

HiveMind

Mercor

OneForma

Remotasks

Toloka

Telus

Top Subject Matter

Computer Vision

Search Algorithms

Geospatial AI

Top Data Types

Image

Video

Audio

Top Task Types

Bounding Box

Mapping

Polygon

Segmentation

Transcription

Evaluation/Rating

Freelancer Overview

Results-driven AI Trainer with 1+ year of experience in data refinement, linguistic annotation, and prompt evaluation. Proven track record improving model accuracy through error analysis, quality assurance, and bias-aware dataset optimization for reliable, high-performance AI systems.

Labeling Experience

Freelance Data Annotator and Prompt Evaluator (Hive Micro)

ImageBounding Box

I performed bounding box annotation, image segmentation, and object detection labeling on diverse computer vision datasets. Additional tasks included prompt quality evaluations and RLHF feedback to optimize LLM responses and improve data quality. These efforts were essential to generating high-quality datasets for AI training and evaluation. • Executed precise object detection labeling for advanced vision tasks • Provided RLHF-based feedback to enhance LLM output • Rated prompt and response pairs for quality control • Used Hive Micro’s platform for efficient data annotation

I performed bounding box annotation, image segmentation, and object detection labeling on diverse computer vision datasets. Additional tasks included prompt quality evaluations and RLHF feedback to optimize LLM responses and improve data quality. These efforts were essential to generating high-quality datasets for AI training and evaluation. • Executed precise object detection labeling for advanced vision tasks • Provided RLHF-based feedback to enhance LLM output • Rated prompt and response pairs for quality control • Used Hive Micro’s platform for efficient data annotation

2022 - 2023

Linguistic Annotator and Localizer, Project LightHouse (Crowdgen)

Text

I contributed to multilingual AI training datasets through English-to-Bengali and English-to-Hindi translation and localization. My work focused on linguistic quality assessment and cultural adaptation to ensure accurate, fluent output for NLP and machine translation models. I managed terminology and performed post-editing aligned with quality standards. • Developed machine translation and NLP datasets with a focus on accuracy and fluency • Assessed translation quality using robust evaluation criteria • Localized content for cultural and linguistic alignment • Applied terminology management for consistent dataset output

I contributed to multilingual AI training datasets through English-to-Bengali and English-to-Hindi translation and localization. My work focused on linguistic quality assessment and cultural adaptation to ensure accurate, fluent output for NLP and machine translation models. I managed terminology and performed post-editing aligned with quality standards. • Developed machine translation and NLP datasets with a focus on accuracy and fluency • Assessed translation quality using robust evaluation criteria • Localized content for cultural and linguistic alignment • Applied terminology management for consistent dataset output

2022 - 2023

Map Data Evaluator, Project Milky Way (Crowdgen)

Mapping

I evaluated map data accuracy, assessed points of interest, and analyzed geographic relevance for mapping algorithms. My responsibilities included quality assurance of spatial datasets and business listings to support geospatial AI model accuracy. I followed detailed guidelines to ensure navigation routes and search results were consistently reliable. • Enhanced map data through thorough evaluation and continuous feedback • Assessed spatial information to improve location-based AI models • Conducted reviews of navigation routes for correctness and efficiency • Ensured business listing data met accuracy standards for mapping purposes

I evaluated map data accuracy, assessed points of interest, and analyzed geographic relevance for mapping algorithms. My responsibilities included quality assurance of spatial datasets and business listings to support geospatial AI model accuracy. I followed detailed guidelines to ensure navigation routes and search results were consistently reliable. • Enhanced map data through thorough evaluation and continuous feedback • Assessed spatial information to improve location-based AI models • Conducted reviews of navigation routes for correctness and efficiency • Ensured business listing data met accuracy standards for mapping purposes

2022 - 2023

Data Labeler and Annotator, Project Whatcom (Appen)

AppenImageBounding Box

I performed detailed image and video annotation, object labeling, and search engine evaluation tasks to increase AI model accuracy. My work involved accurate transcription, data labeling, and query assessment according to established guidelines for optimal model training. This role required consistent quality assurance and error analysis for high-performance results. • Successfully labeled large-scale datasets with high precision for complex computer vision tasks • Collaborated with team members using Appen’s proprietary platform and guidelines • Conducted relevance assessments for search engine results to improve user intent understanding • Implemented prompt evaluation to refine algorithm outputs

I performed detailed image and video annotation, object labeling, and search engine evaluation tasks to increase AI model accuracy. My work involved accurate transcription, data labeling, and query assessment according to established guidelines for optimal model training. This role required consistent quality assurance and error analysis for high-performance results. • Successfully labeled large-scale datasets with high precision for complex computer vision tasks • Collaborated with team members using Appen’s proprietary platform and guidelines • Conducted relevance assessments for search engine results to improve user intent understanding • Implemented prompt evaluation to refine algorithm outputs

2022 - 2023

Education

Work History